OpenAI's Apps SDK sucks.So we built a better way to build ChatGPT apps, fast.You can tap into 800M weekly active users, and early land-grab is everything.200+ developers are already building with our framework.Start with zero boilerplate and get:- Interactive dashboards with real-time data - Form widgets with validation + submission - Data visualizations with charts & graphs - API...
TL;DR OpenAI just shipped its Apps SDK — the real infrastructure behind ChatGPT “apps.” Built on top of the Model Context Protocol (MCP), it lets anyone create interactive UI widgets that talk directly to the LLM. Think: the App Store moment for AI.CORE ARCHITECTURE Every ChatGPT app has two halves:The MCP server — defines tools and logic.The widget — a web app compiled to JS and rendered...
Could you share a bit more on what you mean by “backend-only"? Are you thinking of running logic without any UI part?
Node.js isn’t supported yet, but I’d love to hear your use case!
When you run fastapps create, you get a running service that ChatGPT talks to over MCP, not just a code template. You can extend or deploy it like any other web app.
Building interactive widgets for ChatGPT turned out to be way more complex than I expected. Between the MCP protocol, widget registration, build setup, CSP policies, and resource serving — I kept writing the same boilerplate over and over just to get something basic working.That’s why I built FastApps, an open-source framework that lets you build ChatGPT apps with only two files: a Python...
https://github.com/DooiLabs/Chat.js
Hi! I’ve been experimenting with OpenAI's new apps SDK and found it, rough. The example project is unstructured, full of hardcoded file hashes, and takes hundreds of lines just to get something working.So I built Chat.js — an open-source framework that makes building ChatGPT apps actually pleasant.10-line apps instead of 300+ — just define a component’s name, title, schema, and handler....